5 items
Score, fix and rebuild prompts inside ChatGPT, Claude, Gemini and Copilot. Runs on your device: nothing you type is sent anywhere. Better prompts, right where you type them. Prompt Score sits beside the message box in ChatGPT, Claude, Gemini and Copilot. It scores your prompt out of 10 as you type, then helps you fix it before you press send. FIX IT: QUESTIONS THAT FIT WHAT YOU ASKED Click the score, or press Alt+Shift+G. It reads what kind of request you are making (understanding a topic, an email, a post, a summary, a decision, a plan, code, advice and more) and asks only what that kind needs and your prompt does not already say. A question about history gets asked your level and how deep to go; an email gets asked who it is to and what they should do. Tap a ready answer or type your own, and it goes into your message as a labeled line. A BETTER PROMPT IN ONE CLICK Better prompt rebuilds your whole prompt, ready to send, laid out the way that AI reads best: XML tags for Claude, short headed sections for ChatGPT and Gemini. Every word you wrote stays. Sensible defaults fill what it can, and it tells you what it assumed; only details no default can stand in for, such as who an email is to, become a blank in capitals. Undo puts your message back. OR LET THE AI REWRITE IT Ask wraps your prompt in a short request for the AI you are already chatting with: keep every fact, add what is missing, leave a blank where only you know the detail. No API key, nothing to set up. FIND THE EXPERT PROMPT FOR THE JOB When one of the 669 expert prompts on thinkhard.si fits what you are writing, it links to it. RIGHT-CLICK ANY TEXT Select text on any page and right-click to: improve it as a prompt; check whether an AI answer did what you asked; list the facts to check; see what sounds like AI; clean it up for Word or Slack; check a task for a coding agent. PRIVATE BY DESIGN Everything runs in the page you are on. The extension makes no network requests and collects nothing. There is no account. It stores only its on and off switch. Permissions: storage, for that switch, and the right-click menu. The score uses the same ten checks as the free Prompt Grader on thinkhard.si: what to make; who it is for; background facts; length; format; what to avoid; what good looks like; an example; what to do if unsure; a layout the AI reads well. More free tools for working with AI: https://thinkhard.si/tools/
rating_count is the Chrome Web Store ratings count, not a written-review count.